View Full Version : Breaking into the world of watercooling.
so, i finally made the plunge, and ordered my parts to watercool my rig today! :attn:
not the first time i have built a watercooling loop, but the first time i have built my own! (built a loop for a friend a few years back, he is still running it, problem free!!)
decided that i wanted to get my e8400 running a little bit cooler, so hopefully this will do the trick nicely.
dont have a credit card, so i was forced to order the parts from my local shop, but they seem to have a decent enough selection of watercooling components, and fair prices.
parts i ended up getting are as follows:
Pump - Swiftech MCP655
Rad - Swiftech MCR320
Reservoir - Swiftech MCRES-MICRO Rev.2
CPU Block - Swiftech Apogee GT
Im gonna be going with the black Primochill Primoflex Pro LRT 1/2ID tubing, and plan on running the loop Pump -> CPU -> Rad -> Reservoir.
what are your guy's thoughts on the anti-kink coils? it is going inside a 932 HAF, so i dont think i'll have to worry about too tight of bends in the tubing.
also, for coolant, i think i would rather go with a non-conductive fluid, just me being a little freaked out about sticking water so close to my computer, I was thinking about going with Primochill PC Pure Extreme Water Replacement, anyone else use this, or have any thoughts on it?
also, aftermarket tops for the 655, are they worth it, i really do prefer the look of the EK D5 X-Top Rev.2, and it seems to boast superior performance numbers, or is that something i should not bother with?
also, how big of a difference are the fans i have on the rad going to make?
i was thinking i would start with 3 fans (45 CFM) pulling air through the rad, out the top of the case, perhaps adding another 3 to the other side in push/pull configuration if necessary
GMdoubleG
08-20-09, 05:44 AM
Why are you going with the Apogee GT? Price? It's a pretty dated CPU block. I don't think the aftermarket top for your D5 will help you much since you are only putting one water block on the loop. The quality of the fans are going to determine their longevity and how quiet they run. Just make sure that you are pulling air from outside the case (cold) through the rad.
hokiealumnus
08-20-09, 08:17 AM
I doubt you’ll need anti-kink coils in a HAF. I’ve got some relatively tight bends in my RC690 and have no kinks. If the tubing is good (and Primochill LRT is excellent), you shouldn’t have to worry about it. Unless you want the extra color of course.
For liquid, water is non-conductive. It’s minerals in water that are conductive. Get distilled and it will be as non-conductive as any consumer-grade liquid you can buy. There are truly non-conductive fluids that are not just marketing gimmicks, but the average person cannot afford them.
In my (and many others’) humble opinion, use distilled water + PT Nuke or a silver kill coil. You can get PT Nuke at any number of online dealers (Sidewinder, Petra’s, Jab-Tech, etc); you can also get a Paypal account without a credit card AFAIK, you just need a bank account.
Here’s a D5 pump top shootout (http://www.skinneelabs.com/Pumps/D5/D5-Tops-main.html) to help guide you if you choose to go with an aftermarket top.
Why are you going with the Apogee GT? Price? It's a pretty dated CPU block. I don't think the aftermarket top for your D5 will help you much since you are only putting one water block on the loop. The quality of the fans are going to determine their longevity and how quiet they run. Just make sure that you are pulling air from outside the case (cold) through the rad.
i went with the GT, because sadly, its the best cpu waterblock that my local shop will bring in for me. if i find that temperatures arent quite what im hoping for, i'll be getting myself a prepaid credit card, and ordering a better one for myself online.
and the only reason i was thinking about an aftermarket top for the D5 is because i was planning on getting myself a 4890(hopefully a second one too in the future, currently only $219 for the XFX one at my local shop!), a full coverage block for it, and then possibly the primochill typhoon 3(which may cancel out the ability to use the aftermarket top i think?), and a second rad, to make two loops out of it.
and with the pulling air from outside the case to go through the rad, since im gonna be mounting my rad at the top of my 932 HAF, will that not mess up the whole internal airflow, since it currently pushes air out the top of the case? the fans that i have in the case right now, and plan on mounting to the rad, are the Yate Loon 45CFM, 24DB fans
ricflairw000
08-20-09, 10:59 AM
I doubt you’ll need anti-kink coils in a HAF. I’ve got some relatively tight bends in my RC690 and have no kinks. If the tubing is good (and Primochill LRT is excellent), you shouldn’t have to worry about it. Unless you want the extra color of course.
For liquid, water is non-conductive. It’s minerals in water that are conductive. Get distilled and it will be as non-conductive as any consumer-grade liquid you can buy. There are truly non-conductive fluids that are not just marketing gimmicks, but the average person cannot afford them.
In my (and many others’) humble opinion, use distilled water + PT Nuke or a silver kill coil. You can get PT Nuke at any number of online dealers (Sidewinder, Petra’s, Jab-Tech, etc); you can also get a Paypal account without a credit card AFAIK, you just need a bank account.
Here’s a D5 pump top shootout (http://www.skinneelabs.com/Pumps/D5/D5-Tops-main.html) to help guide you if you choose to go with an aftermarket top.
thanks for the link on the D5 tops
i always wondered about that
hokiealumnus
08-20-09, 11:45 AM
The typhoon 3 is an aftermarket top (and a darn good one at that (http://www.skinneelabs.com/Pumps/D5/TyphoonIII/Primochill-TIII.html)), so no need to worry about one now if you're getting that later.
The typhoon 3 is an aftermarket top (and a darn good one at that (http://www.skinneelabs.com/Pumps/D5/TyphoonIII/Primochill-TIII.html)), so no need to worry about one now if you're getting that later.
kk, thats what i thought it looked like.
and im hoping to get one of those later, they look cool, the dual loop thing seems like a good idea for when i later wanna watercool the future gfx card.
how much performance will i lose by having the air sucked from inside my case, out through the rad, versus having cool air from outside my case blown into the case through the rad?
ricflairw000
08-20-09, 12:23 PM
i think im reading the data wrong but it looks like the stock one is doing the middle of the road on the other pumps and jsut as good as the T3
hokiealumnus
08-20-09, 01:12 PM
i think im reading the data wrong but it looks like the stock one is doing the middle of the road on the other pumps and jsut as good as the T3
Well, you're reading part of it right.
The first part of your sentence is correct, the stock top fits in the middle of the pack. Were I to get an aftermarket top, it would be the Detroit Thermo or the EK top outlet.
For the second part of your sentence, you're definitely reading the data wrong. The TIII is designed to run two loops. Even with uneven restriction in the loops, both of the TIII's individual outlets ran at higher flow rates than the regular tops with the same loops run in series. In short, it trounced other tops when considering loops with a lot of equipment in them.
Ironically, the closest competitor in the series loops was a DDC w/ an XSPC top. Love that little pump!
ricflairw000
08-20-09, 01:24 PM
Well, you're reading part of it right.
The first part of your sentence is correct, the stock top fits in the middle of the pack. Were I to get an aftermarket top, it would be the Detroit Thermo or the EK top outlet.
For the second part of your sentence, you're definitely reading the data wrong. The TIII is designed to run two loops. Even with uneven restriction in the loops, both of the TIII's individual outlets ran at higher flow rates than the regular tops with the same loops run in series. In short, it trounced other tops when considering loops with a lot of equipment in them.
Ironically, the closest competitor in the series loops was a DDC w/ an XSPC top. Love that little pump!
Thanks for the help!!!
yeah thats the worst situation
being partly right haha
OK i thought i would be flowing 2-3 GPM
http://www.skinneelabs.com/Pumps/D5/TyphoonIII/D5-T3_single-loop_s5.jpg
so looking at this i was thinking its only a marginal difference
hokiealumnus
08-20-09, 02:07 PM
Looking at that, you're correct...but that's not what it is designed for. Set up a series loop like this: (Res > Pump > Rad > CPU > Rad > GPU x2 > Res), then compare that to two loops off of the TIII ((TIII > Rad > CPU > TIII) + (TIII > Rad > GPU x2 > TIII)) and the TIII will rule the day.
ricflairw000
08-20-09, 02:12 PM
Looking at that, you're correct...but that's not what it is designed for. Set up a series loop like this: (Res > Pump > Rad > CPU > Rad > GPU x2 > Res), then compare that to two loops off of the TIII ((TIII > Rad > CPU > TIII) + (TIII > Rad > GPU x2 > TIII)) and the TIII will rule the day.
OH OK now all the pins lined up and im getting it. wow two loops off one pump that would save $$$ and space! that would be great!
kk, got it all hooked up and leak testing atm.
to get all the air bubbles out of the system, im gonna have to make the res the highest point in the loop, correct?
will my d5 pump be ok running with the inlet pointing up, and the outlet horizontal?
hokiealumnus
08-20-09, 09:12 PM
It doesn't have to be the highest point in the system, no. The bubbles will naturally bleed out as the water cycles through. You'll be rid of the majority within a couple hours. A few stragglers will stick around for a day or two. Nothing to worry about.
Spawn (I think?) runs his D5s like that and they're fine.
i've had mine leak testing for the past two hours now, and due to the black tubing, the only place i can see the coolant is in the res, and its still milky white with air bubbles.
however, i have needed to top it up a couple times, and seeing as how there has been no leaking yet, this tells me that it must just be air working its way out of the system, correct?
also, will the air bleed out easier if i have the pump speed turned down, as it is currently at speed 5, and sorta looks like its probably just recirculating a whole bunch of the air
hokiealumnus
08-20-09, 09:32 PM
That's correct, it's air working its way out of the system. I'm not sure about bleeding with a D5, never had one so I don't know if changing speeds would help. I do know the MCP355 has only one speed - super fast. I'm spoiled with the XSPC res top though...it bleeds at lightening speed. One thing you can do is tap the side of your rad with a screwdriver. Give it a few good licks (not enough to dent it of course)...it'll help get the stragglers out of there.
My system bled 90% of the way there in a few minutes. I started with the D5 on 1 and bumped it a few times, and turned the speed up bit by bit. I was surprised how quick it was actually, given what I had read!
Also, helps to tilt your case side to side gently. That freed up a few big bubbles for me.
kk, i turned the pump speed down to one, and now the coolant in my reservoir only has a few barely noticeable bubbles floating around in it.
just turned the pump speed back up, and the water is still coming through the res almost crystal clear.
gonna leave it for a little while longer for leak testing, then tomorrow, i'll go buy some compressed air, give it a good dusting, and take pictures!
ricflairw000
08-21-09, 10:36 AM
kk, got it all hooked up and leak testing atm.
to get all the air bubbles out of the system, im gonna have to make the res the highest point in the loop, correct?
will my d5 pump be ok running with the inlet pointing up, and the outlet horizontal?
i called up swifttech they said as long as the inlet is not pointing down you are fine
markp1989
08-21-09, 11:18 AM
That's correct, it's air working its way out of the system. I'm not sure about bleeding with a D5, never had one so I don't know if changing speeds would help. I do know the MCP355 has only one speed - super fast. I'm spoiled with the XSPC res top though...it bleeds at lightening speed. One thing you can do is tap the side of your rad with a screwdriver. Give it a few good licks (not enough to dent it of course)...it'll help get the stragglers out of there.
i have 1, it does bleed very fast :D i got most of the bubbles out in the first few minutes, there were a few small stuborn ones, but they are gone now :D
i read a tip some where , of putting somthing U shaped between the intake and output of the res/pump. to cause turbulance in the res, to stop the bubbles getting sucked back in, i tried it using a straw, that was cut and it did work very well (remove it after bleeding lol)
kk, i got the system up and running, no leaks after leak testing overnight, however, my idle temperatures are 10 C warmer than they were with my aircooling setup.
i figured maybe i goofed on the thermal paste, so i took off the cpu block, and reapplied the AS5, but still my idle temperatures are 10 C warmer.
now, i dont have the fans in push/pull configuration on the rad, i just got some 41CFM fans pulling air through the rad currently, ran outa money for the three high pressure 133cfm fans, and fan controller that i wanted to get, so i could run a push/pull, so thats gonna have to wait until my next paycheck, but could that be my issue?
or my other thought, is the CPU block just too dated, and cant handle the heat of my CPU?
scorchingice
08-21-09, 01:56 PM
Straggling in Late once again on a post :) (Hey I got a new boat that I am putting a new floor in ATM (Well new for me its actually a Very old boat that I took in to restore. It's a classic. Some people do car restoration. I do boats and computers. But I enjoy water with both :) )
you can also get a Paypal account without a credit card AFAIK, you just need a bank account.
Statement is true. You do not even have to have a checking account. Just a savings that allows electronic deposits. (This is the way I did it.)
i went with the GT, because sadly, its the best cpu waterblock that my local shop will bring in for me. if i find that temperatures arent quite what im hoping for, i'll be getting myself a prepaid credit card, and ordering a better one for myself online.
BTW all it takes to get the prepaid is about $8 and a trip to Walmart for their Walmart money card you actually just pull from a shelf, Buy and load with cash :)
NOW: On to the main issue:
If your local shop will not get what you want, Buy it online. In the "Old days" Buying online was risky, But with reputable dealers like Petras, Newegg, FrozenCPU, etc, There is no reason to cut corners because your Local shop isn't up to par.
Anyhow people are correct. The Apogee GT is outdated and not worth getting IMO, Take the time to review blocks, and get the one best for you. Also take into consideration whether your CPU IS Lapped currently or if you are planing to do so in the future. The Apogee GT will be worse than air cooling if you Lap your CPU, the GTZ isn't as bad on Lapped CPU's cause of the new mounting system. But others are better.
Some nice CPU Water block reviews I have found are :
If using an UNLAPPED Stock Processor:
http://vapor.skinneelabs.com/i7/Round1/Round1i7.html
If you have a Lapped or are planning to Lap your CPU: (Which IMO helps SO Much I dropped 18c under full load)
here is a nice review with lapped processors. (Note comment #11)
http://www.xtremesystems.org/forums/showthread.php?t=224976&highlight=heatkiller
However it seems you have your items, and have it up and running.
10c Hotter than the Aircooling. something is wrong. Air in the line, or (Is your CPU Lapped?) if not maybe check the mounting.
I want to clarify for myself. What is your EXACT Loop. What is being cooled with this setup?
If you have just your CPU in this setup with a 3x120 rad and no other blocks, 41 CFM fans should still get you better cooling than you had on air as long as you have all 3 installed and running and your CPU is not lapped. Even with the after market Thermalright Ultra-120 eXtreme, Any watercooling setup should never run 10c hotter.
Just so you do not think I am a rookie blowing smoke (I know how to research)
http://www.evga.com/forums/tm.asp?m=100819358
Note the comment by XLAR8 Post #3 to be exact. He is correct that radiator works better with lower CFM fans actualy, as by design.
Just Note, THAT is your exact CPU in his comment. (Yes he is only at 4GhZ not 4.5GhZ) But still its a nice find for your situation.
hokiealumnus
08-21-09, 02:02 PM
Heh, oops...you linked to the same review twice.
scorchingice
08-21-09, 02:27 PM
Heh, oops...you linked to the same review twice.
Thank you Hokie. Corrected.
BTW I am So Happy I lapped my CPU and cooler. 18c lower temps under full load.
Still haven't jumped to water yet, but actually happy about it. I have changed my mind for the good over 10 times in the past week or 2. And I would have bought a Water block that would have been hurt by using a Lapped CPU.
BTW here is YET another that shows Lapping a CPU can cause adverse effects with Water blocks.
http://www.xtremesystems.org/forums/showthread.php?t=130142
NOTE the comment about :
Wow!! The Apogee with the bowed base outperformed the FuZion by 3.2c!
From about 5 hours searching the Web. I have come to the conclusion that with NO DOUBT:
If you have a Lapped CPU, Lap your Block as well. And check your mounting. (If your block has a mounting setup with a Fixed radius for tightening it down) that means it was designed and setup for a STOCK CPU, NOT a Lapped one and THIS will be a major factor in it performing like crap. It will never tighten down enough to allow proper cooling.
So, IMO Lapping is well worth it, IF you have the right block to support it, otherwise Stock is better.
5 hours reasearching this. Glad I have. anyone with support/debate is welcome to bring it on...
Anyhow back to the matter at hand. YOUR Radiator Specs from Swiftech.
http://www.swiftnets.com/products/MCR320-QP.asp
Note the part of being Optimized for 120mm fans with low to medium air output.
The charts also show with various fan outputs and the details of that.
Your 41cfm fans are NOT the issue. Will higher CFM fans help more, sure most likely they will.
HOWEVER, those fans are NOT causing the issue that you are running at 10c higher than air cooling.
Something else is the issue. Slow fans work fine for this radiator Therefore, NOT the issue for that type of temp severity.
yeah, i have an unlapped processor, im not quite brave enough to attempt lapping right now, maybe i'll attempt lapping an old rig or two of mine first, see how that goes.
and my local shop is actually a pretty good one, most of the time, as i live about 5 minutes away from an (NCIX (http://www.ncix.com) location.
and its only 2 bucks for a prepaid credit card if you goto money mart, or one of the banks that has it, i was just too lazy to go wait in line to get one.
i've redone the thermal paste 3 times, just incase i messed it up, it has looked just fine each time, im using AS5 paste.
i have black tubing, but the air in the reservoir is totally bubble free, so i dont think there is any air bubbles, i spent quite some time tapping on the rad when i was bleeding the system through.
and the only thing this is cooling is my CPU, with the loop going Res -> Pump -> Rad -> CPU -> Res
and the system its in is exactly whats in my sig
thorilan
08-21-09, 05:03 PM
could you elaborate on lapped cpu ;)
you mean heat spreader ?
scorchingice
08-21-09, 07:30 PM
could you elaborate on lapped cpu ;)
you mean heat spreader ?
No I mean Lapped CPU. Lapped CPU is the term used for when you basically take your CPU and make it a TRUE flat surface by Sanding it down with many grades of sandpaper on a true flat surface. That way is is more in the lines of flat and less in the lines of being concave/convex. This gives the CPU the flattest available surface for the heatsink to cool. And the most area for it to cool with no air gaps/less thermal compound. True metal to metal achieves best thermal heat dispersing.
However on SOME Waterblocks/Cooling Blocks, the mounting surfaces were made to best suit stock CPU's due to the fact they have a "Set range of motion"
They will only crank down on a CPU so far (Which is usually the best for Stock CPU's but by lapping them, you lose a little surface area so the heatsink can no longer firmly secure itself to the top of the CPU.
I lowered my CPU temps 18c Under full load on air cooling by Lapping my Intel e6850 and my Zalman 9700 CNPS LED cooler. So yes it was well worth it.
CPU Lapping Guide I used to Lap my CPU. Very nice guide in Youtube format.
It is 2 parts:
Part #1
http://www.youtube.com/watch?v=sVXuZTuoEuE&feature=channel_page
Part #2
http://www.youtube.com/watch?v=5L1s1-nXj0o&feature=related
Conumdrum
08-21-09, 10:45 PM
Some waterblocks still are bowed, can't remember the famous one (V1?) that did better with a thicker o-ring inside the block, thus bowing the block. It worked. I put much more pressure on the CPU exactly where the Cores were. It sqished out the TIM, so there was less TIM (which contrary to belief is an insulator in comparison to a perfect metal to metal contact) so the heat transfer was better.
Some CPU heatsinks, mainly air can benefit from lapping due to manufacturing defects. Some CPU's need lapping.
I don't lap till I place some ketchup on a piece of glass, place the suspect CPU or heatsink on it, push hard, pick up the glass and see how much ketchup is left with the CPU or HS still pushed hard against the glass.
Blindly lapping cuz Joe did it isn't smart. lap if ya need to. And 600 is more than enough.
Spawn-Inc
08-21-09, 11:26 PM
No I mean Lapped CPU.
a lapped cpu is also know as a heatspreader, or integrated heat spreader (IHS) for its full name. it's the metal cap/lid on your cpu and what is often uneven and not very flat.
scorchingice
08-22-09, 12:08 PM
Some CPU heatsinks, mainly air can benefit from lapping due to manufacturing defects. Some CPU's need lapping.
All CPU Heatsinks can benefit from lapping (whether done at the manufacturer or if you do it yourself. (Which if you do it yourself correctly is going to be better than what they do simply because they are mass manufacturing them.
Blindly lapping cuz Joe did it isn't smart. lap if ya need to. And 600 is more than enough.
You are not going to harm anything by lapping as long as you do it correctly and do not over lap your cpu. And using one sheet of 600 is not enough, you are not going to get a nice flat smooth finish with it. And that smooth mirror finish helps with more than just a cool reflection you will never see. It provides less places the heat sink compound must fill for true smooth surface to surface contact.
could you elaborate on lapped cpu ;)
you mean heat spreader ?
a lapped cpu is also know as a heatspreader, or integrated heat spreader (IHS) for its full name. it's the metal cap/lid on your cpu and what is often uneven and not very flat.
Yes it can be called a heat spreader as shown in Wiki, But when I hold my CPU or order a CPU, I don't say Give me a CPU with built in heat spreader. And I don't say "Wow this is a really nice CPU/IHS combo I got in my hand. I say CPU or Processor.
People knew what I meant. Shakes head.
Back to the matter at hand.
Sorry dyckah I can't help beyond what I've asked. With your loop, you should get better temps than air. It may be that GT Block, It seems it doesn't do well with that CPU especially over clocked so high.
Anyhow I hope the pro's here can give you better advice to get those temps down with the info you have stated here.
Spawn-Inc
08-22-09, 06:18 PM
Yes it can be called a heat spreader as shown in Wiki, But when I hold my CPU or order a CPU, I don't say Give me a CPU with built in heat spreader. And I don't say "Wow this is a really nice CPU/IHS combo I got in my hand. I say CPU or Processor.
yes i didn't mean to come across like that. i was just sticking up for thorilan, he knows what a lapped cpu is, or i should hope so :)
kk, so i figure that im going to see about a new cpu block, see if that helps my temperatures get down like i want them.
whats the difference between the Heatkiller CPU Rev3.0, and the Heatkiller CPU Rev3.0 LT?
from what i can see, one of them just has a copper finish on the top, and one has a black finish?
Spawn-Inc
08-24-09, 01:28 AM
correct one is copper and the other is copper and a plastic top. the lc is for low cost. both work great though.
correct one is copper and the other is copper and a plastic top. the lc is for low cost. both work great though.
kk.
think i would be better off with one of those, the Apogee GTZ, or the D-tek fuzion v2?
From a pure performance point of view, yes, you are better off with a Heat Killer. However, I recently got an Apogee GTZ -- mainly cause I figured the mounting system was excellent, and was one less thing that could potentially go wrong. For me, the couple of degrees worse temps were worth a dead easy mount.
YMMV of course.
thorilan
08-24-09, 03:48 AM
my point was its a heat spreader.
the reason i point that out is because some of us de lid our CPUs and i have even known some people to try and lap a core ( dont ask me why)
havent de lidded a CPU in a while . maybe i will consider the 9850 amd a candidate.
Hipcrostino
08-24-09, 04:32 AM
I have to agree with thorilan here. I remember the days of lapping the core, but hey we hardly have IHS's in them days.
I would also like to point out that lapping to a mirror finished has been conclusivly prooved to have no benifit over a not so mirror finish. I mean, you want it smooth, and will be able to see yourself in it, but you won't be able to start fires with it or pop pimples in it. The key is smooth and flat.
@ dyckah - I wouldn't both with a new cpu block quite so soon. If your temps are higher than air at idle, the most important thing to remember is that idle temps don't matter. Load temps are what counts. I would suggest that first up you make sure you block is tight on the cpu. My Apogee GTX is so tight the board is bowing a tad. You don't need to go that tight(and probably shouldn't), but ive found that pressing the HS on the cpu with more pressure can bring down temps by 10s of degrees. So make sure its tight. Test your load temps, see if they are better than air, and if your happy with them. If your still not, then upgrade the block. I just don't think they suck as much as everyone is making out. Sure they are old, but a new block won't net you 10c in load temps unless something other than just the old block was incorrect.
wow that was longer than expected...!
I have to agree with thorilan here. I remember the days of lapping the core, but hey we hardly have IHS's in them days.
I would also like to point out that lapping to a mirror finished has been conclusivly prooved to have no benifit over a not so mirror finish. I mean, you want it smooth, and will be able to see yourself in it, but you won't be able to start fires with it or pop pimples in it. The key is smooth and flat.
@ dyckah - I wouldn't both with a new cpu block quite so soon. If your temps are higher than air at idle, the most important thing to remember is that idle temps don't matter. Load temps are what counts. I would suggest that first up you make sure you block is tight on the cpu. My Apogee GTX is so tight the board is bowing a tad. You don't need to go that tight(and probably shouldn't), but ive found that pressing the HS on the cpu with more pressure can bring down temps by 10s of degrees. So make sure its tight. Test your load temps, see if they are better than air, and if your happy with them. If your still not, then upgrade the block. I just don't think they suck as much as everyone is making out. Sure they are old, but a new block won't net you 10c in load temps unless something other than just the old block was incorrect.
wow that was longer than expected...!
kk, if i was to tighten up the block onto the cpu, should i reapply the thermal paste, and then tighten up, or just tighten the thing up as is?
i last put the cpu block on on friday
edit::
so i just tightened up the block as is, and due to the high room temperature, and still not getting temp's i was comfortable with while loading the cpu, i dropped the cpu multiplier down to 8, and the clock down to 4.0Ghz, i got a max temp of 75 C while using IntelBurnTest, with prime95 small ftt, it didnt go over 62 C!
and some pictures of my loop!
http://img18.imageshack.us/img18/3633/2408091717.jpg
http://img43.imageshack.us/img43/1476/2408091718.jpg
http://img6.imageshack.us/img6/731/2408091719.jpg
http://img6.imageshack.us/img6/8391/2408091720.jpg
http://img8.imageshack.us/img8/2330/2408091722.jpg
http://img43.imageshack.us/img43/4751/2408091723.jpg
http://img43.imageshack.us/img43/5562/2408091724.jpg
Looks good! I'd personally prefer less bling (LEDs), but each to their own :)
Daddyjaxx
08-24-09, 09:13 PM
Looks good...but I'm blind now. :)
hokiealumnus
08-24-09, 09:51 PM
Looks great; nice work!
Nice and tidy! Good job :thup:
Looks good...but I'm blind now. :)
lol.
and here was me thinking i wanted to add another couple cathode ray tubes, mod the stock red 230mm fan so that its blue, put that back in, and add three more LED fan's to the bottom of the rad :attn:
Conumdrum
08-24-09, 10:57 PM
OMG, you need to go to XS and look up the "Bling King Darth Beevis" His rigs glow with with fortitude and eye candy all the way. Nice guy in RL too. Respected in the WC community.
I personally don't get the bling thing... My computer should be fast and quiet (ie intrusive in every way). If I could fit all my stuff into a little shuttle-type case, I would!
Hipcrostino
08-25-09, 12:01 AM
It dose look good. Very neat!
75 is quite high i think. My system was getting 48c loaded last night at 4ghz. ambient is ~24c. What voltage are you using on the chip? Mines about 1.200v when loaded in CPUZ. 1.225 in bios. The IntelBurn test will load the cpu much more than anything else. It really puts on the heat. I should add that it was 48c under the program too. The GT and GTX are the same except the inlet is further away from the outlet on the GTX. Not sure about the GTZ. It made 2-3 degree difference, do that dosn't account for the high temps your getting. The only think I can think of now is that either you have air in the system still, or your IHS is not flat. It might be time to lap it and your block.
It dose look good. Very neat!
75 is quite high i think. My system was getting 48c loaded last night at 4ghz. ambient is ~24c. What voltage are you using on the chip? Mines about 1.200v when loaded in CPUZ. 1.225 in bios. The IntelBurn test will load the cpu much more than anything else. It really puts on the heat. I should add that it was 48c under the program too. The GT and GTX are the same except the inlet is further away from the outlet on the GTX. Not sure about the GTZ. It made 2-3 degree difference, do that dosn't account for the high temps your getting. The only think I can think of now is that either you have air in the system still, or your IHS is not flat. It might be time to lap it and your block.
1.232V in CPUZ, not exactly sure what i set it at in bios, just a little bit higher.
abient temp is around 27c, idle is at 38c.
i dont think i have air in my system, i sat for like 20 minutes, tapping away at my rad, trying to get all the air bubbles out, and the fluid in my reservoir is totally bubble free.
kinda freaked out about the whole lapping process, having never done that before, and not wanting to wreck my processor, although i suppose the warrenty is already void on it, since i have been overclocking it..
hokiealumnus
08-25-09, 07:32 AM
Since I've got the same CPU, just for reference, I max out at ~55C with ambient at 27C running Prime95. Folding@Home at the same ambient gets to 49-50C. Drop the ambient to ~22-23C, my folding temps drop to a cool 45-46c. Never tried IBT; couldn't get it to run on my last system and never downloaded it for this one.
ricflairw000
08-25-09, 10:59 AM
i like the glass with steel rebarb in it :D
Since I've got the same CPU, just for reference, I max out at ~55C with ambient at 27C running Prime95. Folding@Home at the same ambient gets to 49-50C. Drop the ambient to ~22-23C, my folding temps drop to a cool 45-46c. Never tried IBT; couldn't get it to run on my last system and never downloaded it for this one.
kk, well, im only running like 5-6c hotter than you while running Prime95, so i suppose that i've got a decent starting place.
and the reading i've done into my block is telling me that i will get better temperatures if i swap in the thicker o-ring, so i'll probably test that out a little later today/tomorrow, whenever my cable sleeving kit, more ram(G.SKILL F2-8000CL5D-4GBPI Pi PC2-8000 4GB 2X2GB), and typhoon 3 come in(brother decided to surprise me with an early xmas present, and asked me what i wanted, so more computer stuff was the first thing to come to mind!)
oh, so today i added another 3 fans to my rad, to see what sorta temps i would get then, and idle temperatures didnt change, but my intelburntest temps dropped 5c!
and now my case is that much brighter again, lol.
currently i have the fans pushing air from inside my case through the rad, and out the case.
how much better would my temperatures fair if i had the fans pulling air down in through the rad, into the case? and how much would that heat up the rest of my case?
vBulletin® v3.8.7, Copyright ©2000-2012, vBulletin Solutions, Inc.